1. Field of the Invention
The present invention relates to a method for avoiding irregular shutoff of production equipment equipped with a vacuum pump and a system for avoiding irregular shutoff.
2. Description of the Related Art
Using failure prediction techniques for a vacuum pump attached to semiconductor production equipment, a life expectancy prediction may be made for a vacuum pump from a long-term perspective in order to perform scheduled maintenance. However, since a thin film deposition is carried out under a plurality of processing conditions having varying pump loads during normal operation of the production equipment, in cases where film deposition is carried out under process conditions where a pump load is high, there are times when the vacuum pump will suddenly shut off during the film deposition.
If such a situation occurs, in many cases entire product lots in the middle of the film deposition process will be lost. In addition, due to irregular shutoff of the vacuum pump during the film deposition, it is possible that a highly reactive gas and/or a highly toxic gas may be released into the air when the vacuum pump is replaced, deteriorating the work environment and causing problems for employee health maintenance. In particular, in a case of semiconductor production equipment, there are times where a highly toxic gas such as arsine (AsH3), phosphine (PH3), or diborane (B2H6) is used. If the vacuum pump suddenly shutoffs while using these gases, the possibility of serious or even life-threatening accidents happening cannot be ruled out. In the case of the vacuum pump for evacuating such highly toxic gases and/or highly reactive gases, there are situations where special measures, such as replacing the vacuum pump in a draft chamber, must be taken. Therefore, depending on the type of the gas used, more labor hours may become necessary than for an ordinary pump exchange.
In addition, maintenance may be performed after a vacuum pump life expectancy prediction or a prediction of when an irregular shutoff caused by motor current or the temperature of a cooler within the vacuum pump may occur according to experience-based equipment management. However, depending on the gas used, even if these parameters fall within a normal range of values, the vacuum pump may be in trouble. Therefore, whether in a single process or in multiple processes, it is impossible to assess a failure using only these parameters. Furthermore, due to the fact that only pump data is analyzed, in the case where the multiple processes are performed the parameters monitored fluctuate greatly because the pump load varies with each condition, and a life expectancy prediction is impossible.
As described above, in cases corresponding to production of various kinds of industrial products, when the same production equipment is used to manufacture various kinds of industrial products, the required process conditions are diverse, and there is a problem in that it is impossible to determine a universal threshold or criteria for production equipment failure prediction.